Optical communications networks have been deployed for many years. Originally, these networks were generally “point to point” type networks including a transmitter and a receiver connected by an optical fiber. Such networks are relatively easy to construct but deploy many fibers to connect multiple users. As the number of subscribers connected to the network increases, the fiber count also increases rapidly and the expense of deploying and managing many fibers becomes complex and expensive.
A passive optical network (PON) addresses this problem by using a single “trunk” fiber from a transmitting end of the network, such as an optical line terminal (OLT), to a remote branching point, which may be up to 20 km or more. Each subscriber may thus be connected to the network utilizing a much shorter fiber span forming an architecture sometimes referred to as “tree and branch.” One challenge in developing such a PON is utilizing the capacity in the trunk fiber efficiently in order to transmit the maximum possible amount of information on the trunk fiber.
To improve efficiency, PONs have used “time domain multiplexing” by assigning each subscriber on the PON a particular time slot for transmission of its data along the trunk fiber. Each subscriber is allowed to transmit during its assigned time slot, and these slots are synchronized at both the transmitter and receiver such that the receiver knows the time slot (and thus the transmitter) of a transmitted signal. In this way, many transmitters can share the same fiber without fear of multiple transmitters sending data at the same time and confusing the receiver. Standards such as gigabit PON (G-PON) and Ethernet-based PON (E-PON), for example, utilize this time-dependant approach.
Although TDM-PON systems work, the TDM approach is inefficient because the system should allow sufficient time between different transmitter time slots to prevent confusion at the receiving end. Also, noise in this type of system is cumulative across all the transmitters in the PON. To avoid unwanted noise, transmitters other than the one currently transmitting may be turned off and then turned on rapidly when it is time to transmit data, without providing much stabilization time. This “burst mode” transmission makes it challenging to increase data rates in a TDM-PON system.
TDM also does not make efficient use of the bandwidth available on the fiber. Optical fiber has the ability to carry many different signals simultaneously, without interfering, as long as these different signals are carried on different wavelengths. TDM-PON systems utilize only a few wavelengths and therefore do not utilize much of the fundamental bandwidth available on the optical fiber. Similar to radio transmissions utilizing different frequencies to carry different signals, fiber optic communications networks may increase the amount of information carried on a single optical fiber by multiplexing different optical signals on different wavelengths using wavelength division multiplexing (WDM).
In a WDM-PON, a single trunk fiber carries data to and from an optical branching point and the branching point provides a simple routing function by directing signals of different wavelengths to and from individual subscribers. In this case, each subscriber is assigned a particular wavelength on which to send and/or receive data. The WDM-PON thus allows much greater bandwidth because each transmitter is allowed to transmit at a higher data rate and for a longer period of time.
One of the challenges in WDM systems, such as a WDM-PON, is coupling an array of lasers to an optical multiplexer such as an arrayed waveguide grating (AWG). Lens and/or fiber assemblies are sometimes used to provide proper optical coupling efficiency in existing systems; however, these assemblies tend to require a number of parts, which are expensive, complicated, and require space.